Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A construction method and application of a lightweight gesture detection convolutional neural network model

A convolutional neural network and gesture detection technology, applied in the field of computer vision, can solve the problems of limited application of network models, time-consuming calculations, etc., and achieve the effect of occupying less computing resources, less computing, and fewer network layers

Inactive Publication Date: 2019-06-18
HUAZHONG UNIV OF SCI & TECH
View PDF7 Cites 14 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a construction method and application of a lightweight gesture detection convolutional neural network model, which is used to solve the network model of the existing gesture detection convolutional neural network model due to the time-consuming production of training image data and the large amount of calculation. The problem with limited applications

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A construction method and application of a lightweight gesture detection convolutional neural network model
  • A construction method and application of a lightweight gesture detection convolutional neural network model
  • A construction method and application of a lightweight gesture detection convolutional neural network model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0055] A construction method 100 of a lightweight gesture detection convolutional neural network model, such as figure 1 shown, including:

[0056] Step 110, based on the SqueezeNet convolutional neural network architecture and the SSD multi-target detection convolutional neural network architecture, construct a lightweight gesture detection convolutional neural network framework;

[0057] Step 120. Obtain a preset number of gesture pictures and background pictures, and perform image data enhancement and picture synthesis processing on the gesture pictures based on the background pictures to obtain a gesture data set;

[0058] Step 130: Based on the public dataset and the gesture dataset, train the lightweight gesture detection convolutional neural network framework to obtain a lightweight gesture detection convolutional neural network model.

[0059] This embodiment is based on the SqueezeNet convolutional neural network architecture and the SSD multi-target detection convol...

Embodiment 2

[0061] On the basis of Embodiment 1, step 110 includes:

[0062] Build a feature extraction module, the feature extraction module includes: a plurality of first convolution kernels connected in a preset order, a plurality of pooling units and a plurality of lightweight convolution modules of the SqueezeNet convolutional neural network architecture, used to treat Process images for convolution and pooling operations to obtain multiple feature maps at multiple scales;

[0063] Construct a feature matching module, which includes: a priori frame generation unit connected in sequence, a convolution filter and a fusion unit, which are used to predict the detection target in the picture to be processed based on multiple feature maps.

[0064] Preferably, the prior frame generation unit adopts the prior frame generation unit of the SSD multi-target detection convolutional neural network architecture.

[0065] It should be noted that the lightweight convolution module can reduce the a...

Embodiment 3

[0070] On the basis of Embodiment 1 or Embodiment 2, the convolution filter includes a plurality of second convolution kernels connected in sequence; then in the feature extraction module, a plurality of first convolution kernels, a plurality of pooling units, a plurality of The lightweight convolution module of the SqueezeNet convolutional neural network architecture is used to perform convolution and pooling operations on input images in a preset order to obtain multiple feature maps at multiple scales.

[0071] In the feature matching module, the prior frame generation unit adopts the prior frame generation unit of the SSD multi-target detection convolutional neural network architecture, which is used to generate multiple prior frames corresponding to each feature map; the convolution filter is used for A plurality of second convolution kernels perform convolution operations on the area covered by each prior frame to obtain the first prediction information. The first predict...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to a construction method and application of a lightweight gesture detection convolutional neural network model, and the method comprises the steps: constructing a lightweight gesture detection convolutional neural network framework based on a SquezeNet convolutional neural network framework and an SSD multi-target detection convolutional neural network framework; Acquiring agesture picture and a background picture, and performing image data enhancement and picture synthesis processing on the gesture picture based on the background picture to obtain a gesture data set; And based on the public data set and the gesture data set, training a lightweight gesture detection convolutional neural network framework to obtain a lightweight gesture detection convolutional neuralnetwork model. According to the invention, a small amount of gesture data is expanded into the gesture data set containing a large amount of picture data at a high speed; The technical problem that alarge amount of high-quality gesture picture data is difficult to obtain is solved, in addition, by combining the SquezeNet convolutional neural network architecture and the SSD multi-target detectionconvolutional neural network architecture, the constructed lightweight gesture detection convolutional neural network model occupies few computing resources, and can be applied to various detection platforms.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a construction method and application of a lightweight gesture detection convolutional neural network model. Background technique [0002] Gesture operation is a simple and quick way to realize human-computer interaction. Traditional gesture detection algorithms are mostly based on SVM (Support Vector Machine, Support Vector Machine). This method first extracts gesture features in pictures, and then uses SVM classifier to classify them. Various gestures are categorized. Such methods are usually difficult to extract the salient features of various gestures, resulting in poor generalization ability of the trained gesture detection model. [0003] In recent years, as the computing power of hardware devices has been greatly improved, deep convolutional neural network algorithms have also developed rapidly with the growth of hardware resources. The gesture detection algorith...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06N3/04
Inventor 彭刚任振宇
Owner HUAZHONG UNIV OF SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com